Repbase Update (http://www.girinst.org) is a relational database of repetitive elements representing over 8500 families and subfamilies of transposable elements (TEs) from eukaryotes. Each family is identifiable by a unique name, and its annotation includes species of origin, systematic classification, keywords, reference to the scientific literature, names of contributors, brief commentaries, etc. Most of the 5198 repetitive families added to Repbase Update (RU) during the last cycle are either unreported anywhere else, or have been thoroughly revised. Our approach is based on computer-assisted reconstruction and analysis of public DMA sequence data. Original contributions are first reported in an electronic journal named Repbase Reports. RU is used throughout the world by academic and research institutions for basic research and genome annotations. This database became a unique resource for individual research projects of biological and medical importance and for creation of secondary databases. During the next five years RU needs to grow at the rate of ~1000 entries per year to meet the demand created by the genome-sequencing projects. This information will be extracted, reconstructed, analyzed, annotated, classified, indexed and made available to researchers over the internet. In addition we will discover and study diverse superfamilies and conserved repeats, which are of primary interest for cutting-edge research. We propose the following specific aims to meet the challenge: (1) continue detection, reconstruction, annotation and electronic distribution of reference sequences for repetitive families from all sequenced eukaryotic species (2) continue studies of TEs in newly sequenced eukaryotic genomes in collaboration with sequencing consortia (3) continue systematic identification and analysis of new superfamilies and classes of transposable elements (TEs) (4) excavate and characterize remnants of TEs overrepresented in conserved non-coding regions and cis-regulatory modules in eukaryotic genomes (5) organize two conferences devoted to stimulation of Repbase-related research, data submission, data dissemination, training and standardization of repeat nomenclature.